3. Receptores Volumétricos
3.3. Clasificación de Receptores Volumétricos
3.3.2. Receptor no presurizado con absorbedor cerámico
The commercial availability of required starting material dictated which synthetic strategy was to be adopted. In this regard the availability of the commercial
compound 2,4,6-trimethoxyacetophenone 45 was considered a good starting material for the chemistry outlined in scheme 8.
Scheme 8 OMe Me MeO OMe Me •0 NH2NH2.H20 ►-
OMe AcOH/A MeO' OMe
45 46 PPA/A OMe MeO- 48 [O] OMe Me MeO‘ 47 Br(CH2)4Br OMe C02 MeO-
The ‘one-pot method’ of Zhenqi et.al5 9 was followed in which 2,4,6-trimethoxy
acetophenone 45 was treated with hydrazine hydrate in the presence of glacial acetic acid at 110-120°C. The resulting hydrazone 46 without isolation was subjected to heating in PPA at 110-135°C under stirring for 30 min. to give the indazole 47 in a low yield of 20% (lit. yield 65%). The reaction was repeated several times using fresh commercial PPA but the yield could not be improved to match that reported in the literature. The reaction was next performed using PPE instead of PPA but the product formation deteriorated considerably under these conditions. Product isolation from the reaction mixtures involving PPA proved particularly difficult. Although Zhenghi
neutralisation of the aqueous phosphoric acid with a base, it was considered necessary at the end of the reaction to neutralise the aqueous PPA with either solid sodium hydrogen carbonate or concentrated ammonia solution since the indazole product has basic nitrogen atoms which are likely to form salts with the aqueous phosphoric acid. The reaction mixtures were extracted with the recommended solvent, ethyl acetate, as well as other solvents such as dichloromethane and chloroform. Interestingly, it was not possible to isolate any product at all if the reaction mixture was not neutralised as suggested in the literature. The hydrazone 46 was formed in excellent yield and was characterised by ]H-NMR and IR spectroscopy. The *H-NMR spectrum showed the methyl group resonating at higher field (5 1.0 ) compared with
the methyl in the ketone 45 (5 2.20) and the appearance of a broad signal at 5 8.10 which presumably was the amino group. Its IR spectrum showed two sharp peaks at 3260 cm' 1 and 3350 cm’1 indicative of the amino group as well as the absence of the
carbonyl group at 1690 cm' 1 and the appearance of a peak at 1670 cm' 1 indicative of
the imine bond.
In view of the low yield obtained for the 1//-indazole 47 attention was turned to another method of ring closure reported in the literature55. Using this method the ketone 45 was reacted with hydrazine hydrate and ammonium acetate in xylene at 135°C for 24 h. TLC analysis [ethyl acetate: petrol; 1:3 v/v] of the reaction mixture showed a new spot with very little of the starting material left. The xylene was removed by simple distillation and the product mixture was separated by flash chromatography [ethyl acetate: petrol; 1:3 v/v] to yield the product compound identified as the hydrazone 46 ( 89%). This method did not offer any greater advantage over the previous method.
Hydrazines with electron releasing groups such as alkyl hydrazines have been reported to give best results in the ring closure process to form the indazoles. In this regard initial experiments were conducted with methylhydrazine which is
commercially available. The reaction of methylhydrazine with 2,4,6-
trimethoxyacetophenone using catalytic amount of glacial acetic acid in PPA yielded the corresponding 1,3-dimethylindazole albeit in low yield (18%). This was a
OMe Me 2.PPA/ A MeO' Me 57 Me MeO' OMe 45
The next consideration was whether the hydrophobic content of the expected indazole could be increased by using benzylhydrazine instead of hydrazine and if by so doing the molecule would precipitate out better at the end of the reaction. Benzylhydrazine is commercially available as the dihydrochloride or dihydrobromide salt. The other advantage of benzylhydrazine is that the benzyl group can be removed from the anticipated indazole under mild conditions such as by hydrogenolysis involving hydrogen gas and Pd/C catalyst in an alcoholic solvent (scheme 10).
Scheme 10 OMe Me OMe Me PhCH2NHNH2. 2HBr O HN ‘OMe AcOH/EtOH A
MeO' OMe MeO'
58 PPA/ A OMe Me MeO' 47 OMe Me MeOH MeO' Ph
Benzylhydrazine dihydrobromide was reacted with 2,4,6-trimethoxyacetophenone in refluxing ethanol to form the hydrazone which after removal of the solvent was subjected to cyclisation in PPA at 120-135°C. Disappointingly the isolated crude product was shown to consist of a mixture of many compounds and separation of the mixture by flash chromatography yielded the expected product in very low yield (16%). The NMR spectrum showed the benzyl group having its benzylic protons
resonating at 8 4.80 as a result of deshielding by the aromatic ring and the nitrogen
atom and its phenyl group at 5 7.20-7.35. The methyl group in the 3-position resonated at the expected position of 8 2.75. Electrospray mass spectrum of the
molecule showed an [M+H]+ ion at m/z 283 consistent with the molecular structure. We next diverted our attention to oxidation of the methyl group at the 3-position in the 1//-indazole product 47. Successful experience with oxidation of substituted 4- methylquinolines to the corresponding aldehydes in these laboratories using selenium dioxide in moist dioxane encouraged us to try the reagent for the oxidation of the 3-
methylindazole system 47. Scheme 11 OMe CHO SeO Dioxane MeO' 60 OMe Me MeO' 47 Jones acetone OMe MeO' 48
A literature search had revealed that oxidising agents for a methyl group on aromatic systems include Se0 2 in pyridine and KMnCU solution. It was envisaged that the
oxidation of 47 with Se0 2 would produce the aldehyde 60 which on further oxidation
with Jones’s reagent would produce the desired carboxylic acid 48 (scheme 11). However, when the chemistry outlined in scheme 11 was performed in the laboratory, the crude product isolated from the reaction was flash chromatographed [ethyl acetate: petrol; 1:3 ] to give recovered starting material 47 only. This result was surprising because as pointed out earlier oxidation of 4-methylquinolines with Se0 2 in wet
dioxane produced the corresponding aldehydes in reasonable yields (50-60%). A literature search on indazole-3-carboxylic acids revealed a synthesis that involved
reaction of P-acetylphenylhydrazine 61 with chloral hydrate and hydroxylamine hydrochloride in acidic medium to afford an intermediate TV-acetylaminoisonitroso- acetanilide 62 in 80% yield64. Treatment of the intermediate with sulphuric acid afforded indazole-3-carboxylic acid 64 in 77% yield via compound 63 as shown in scheme 12. For this synthesis to be applicable to the alkaloids studied here would require 2,4,6-trimethoxyhydrazine, which is not commercially available.
NHNHCOCH-j 64 Scheme 12 CI3CCH(OH)2 NH,OH.HCl H2S04 h2o A NOH NNHCOCH3 62 h2so4 63 NHCOCH-,
In view of the difficulty in making 2,4,6-trimethoxyphenylhydrazine required for making the indazole-3-carboxylic acid by the route outlined in scheme 1 2, it was
decided to invent another novel strategy for making 4,6-dimethoxyindazole-3-
carboxylic acid 48. The new strategy involved making the trimethoxy keto ester 65 by Friedel-Crafts acylation of 1,3,5-trimethoxybenzene with ethyl oxalylchloride and hydrolysis of the keto ester 6565 to the corresponding keto acid 6 6 65 required for the
synthesis the indazole 48.
OMe OMe OMe
OEt OH
MeO' OMe MeO OMe MeO' OMe
A literature search showed66 a method for the synthesis of an a-ketoester 6 8 from 2,3-
dihydro-l,4-benzodioxin 67 and hydrolysis of the a-ketoester 6 8 to the corresponding
a-ketoacid 69 as shown in scheme 13.
Scheme 13 67 AlClq C lC 0 C 0 2Et CS2 KOH C02Et EtOH H20 co2h
This strategy was adopted for our system as outlined in scheme 14. The literature
Scheme 14 MeO' OMe OMe MeO' 48 OMe O MeO' OMe 65 KOH/EtOH OMe O H+ /A MeO' OMe 66
procedure66 for making a-ketoester 6 8 used a small amount of anhydrous AICI3 but
we found that a large amount of anhydrous AICI3 was essential for obtaining good
yields of the a-ketoester 65. It was also found that the hazardous solvent CS2 could be
substituted with dry DCM without any difference in yield of the product 65. The a- ketoester 65 was purified by flash chromatography and obtained as a white crystalline solid in excellent yield (94-97%). The a-ketoester 65 was spectroscopically
characterised. Its IR spectrum displayed two carbonyl frequenceies at 1725 and 1678 cm' 1 whilst its proton magnetic resonance spectrum showed a typical ethyl pattern of a
singlet at 8 3.79 and the third methoxy group which is Para to the ketoester functional
group resonated at slightly lower field at 3.84 ppm. The two aromatic protons showed as a singlet at 6.08 ppm. The electron impact mass spectrum of the compound gave the molecular ion peak at m/z 268 whilst the fragmentation ions at m/z 195 and m/z
168 corresponded to the loss of ethoxycarbonyl group and carbonylethoxycarbonyl group respectively. Hydrolysis of the a-ketoester 65 in a mixture (7:3) of ethanol and water produced, after acidification with 6M HC1 acid, the a-ketoacid 6 6 as a grey-
white solid in excellent yield (81-93%). The a-ketoacid 6 6 was readily soluble in
water presumably due to its enhanced acidity due to the a-carbonyl group and hence greater dissociation in an aqueous medium. The IR spectrum of the a-ketoacid 6 6
showed typical features such as broad hydrogen bonded OH group at 2789-2690 cm"1
and two carbonyl absorptions at 1757 and 1724. !H NMR spectrum of the a-ketoacid
6 6 measured in deuterium oxide was very simple as expected. It showed the
resonance of the three methoxy protons at 8 3.82 whilst the two aromatic protons
resonated at 8 6.15. The carboxylic OH group could not be detected at lower field due
to the fact that it underwent exchange with D2O and as a result, however, a large
singlet at 8 4.78 was observed which most probably was due to HDO. The
electrospray mass spectrum displayed the molecular ion at m/z 141 corresponding to (M+l).
It was decided to react the a-ketoacid 6 6 with hydrazine hydrate in water containing a
catalytic amount of acetic acid since both substrates were readily soluble in water.
Scheme 15
H20 /AcOH
The refluxed mixture was evaporated and the residue heated with PPA in order to induce intramolecular cyclisation. The isolated product, obtained in low yield, was examined by proton magnetic resonance spectroscopy in D2O which showed a singlet
that integrated to two protons at 8 1.90 and was assigned to an NH2 group. The two
field at 6 3.77 compared to the starting a-ketoacid 6 6 in which the three methoxy
groups resonated at 5 3.82. The remaining methoxy group Para to the hydrazone functional group resonated at slightly lower field at 8 3.85. The two aromatic protons
resonated as a singlet at slightly lower field at 8 6.35 compared to the a-ketoacid 6 6
in which the aromatic protons resonated at 8 6.15. The electrospray mass spectrum
gave an ion at m/z 255 that corresponded to M+l for the hydrazone 70. The cyclised indazole 48 was not formed in this reaction.
When the reaction of the a-ketoester 65 was performed with hydrazine hydrate in dry DMF in the presence of catalytic amount of glacial acetic acid at 120-130°C the molecule underwent fragmentation to produce the isolated reaction product as 1,3,5-
trimethoxybenzene according to the proton magnetic resonance spectrum of the compound. Similarly, when the was a-ketoester 65 and hydrazine hydrate were heated in o-xylene containing a small amount of tosic acid, the crude reaction product on examination by ]H NMR spectroscopy showed the absence of an ethyl group and indicated the formation of 1,3,5-trimethoxybenzene. There was no evidence for the formation of hydrazone 71 or the indazole 72.
Scheme 16
OMe xylene/AcOH M e0-
In previous experiments the reaction of the a-ketoacid 6 6 with benzylhydrazine
dihydrobromide in o-xylene in the presence of sodium acetate at reflux temperature for 72 h was investigated. This reaction yielded a complex mixture of products which were separated by flash chromatography to yield five fractions that were examined by ]H NMR spectroscopy. Fractions 4 was the cleanest spectrum indicative of
benzylhydrazone formation and fractions five and six although they had elements of the correct resonances for the benzylhydrazone were actually less reliable. Repeating the experiment in dry DMF containing ammonium acetate at 100-120°C the reaction
mixture was once again quite complex and separation of the various products by flash chromatography generated five fractions which were examined by *H NMR
spectroscopy. Fraction four was identified as 1,3,5-trimethoxybenzene whilst fraction five was interpreted to be benzylhydrazine. The other remaining fractions were difficult to interpret and make conclusions.
The a-ketoester 74 from 3,5-dimethoxytoluene was also synthesised by reaction with ethyl Oxalyl chloride under similar reaction conditions that were used for making the a-ketoester 65. The 3,5-dimethoxytoluene 73 itself was made in 60% yield from the commercially available 3,5-dihydroxytoluene by alkylation with dimethyl sulphate using the Williamson ether synthesis. Treatment of 3,5-dimethoxytoluene 73 with ethyl Oxalylchloride in anhydrous CS2 in the presence of AICI3 produced after
purification by silica column chromatography the desired compound 73 in 60% yield as a solid compound. The ]H NMR spectrum of the compound showed the ethyl resonance at 5 1.40 and 6 4.35 whilst the methoxy groups resonated in the expected
positions of 8 3.80 and 8 3.85. The methyl group was observed to resonate at 8 2.45.
The electro ionisation mass spectrum gave the correct molecular ion at m/z 252 (8%)
and fragmentation pattern that was consistent with the correct structure. Scheme 17 CIC0C02Et NaOH l.KOH EtOH/H20, 75
Hydrolysis66 of the a-ketoester was carried out using the literature conditions of dilute
aqueous ethanolic KOH which produced the a-ketoacid 7565 in 90% yield. The
1754 cm’1. Its !H NMR spectrum recorded in deuterium oxide showed a three proton singlet at 2.35 ppm for the methyl group and a six proton singlet at 3.78 ppm for the two methoxy groups. The H-3 aromatic proton resonated at 6.10 ppm whilst the H-5
aromatic proton resonated slightly lower field at 6.15 ppm.
The construction of the six membered ring on the indazole nitrogen was addressed next atoms and for this commercially available indazole was chosen as the model system for the investigation. Earlier work in these laboratories had shown that if the reaction between indazole and 1,4-dibromobutane was carried out in DMF under heating, the product of the reaction was the elimination product 76 (scheme 18). It appeared that in DMF, a polar solvent, the E2 reaction was favoured rather than the desired Sn2 reaction. An alternative solvent for the reaction was considered and it was
decided to use 1,4-dioxane. Heating an equimolar solution of indazole and 1,4- dibromobutane in 1,4-dioxane at 150°C produced the iminium bromide 78 in 64% yield after trituration with dry ether and drying the product in the oven at 60°C.
Scheme 18 Br(CH2)4Br DMF A indazole Dioxane A Br(CH2)4Br
y
The compound was identified by H NMR spectroscopy and mass spectrometry. The electron ionisation mass spectrum of the compound showed two nearly equal isotopic molecular ions at 252 (Br79) and 254 (Br81) (18%) and fragment ions at m/z 173 (53%) corresponding to the loss of a bromine atom. The *H NMR spectrum recorded in deuterochloroform showed a broad singlet at 8 1.90 which was probably water. A
four proton multiplet at 8 2.35 was due to the two internal methylene groups while the
other two methylene protons adjacent to the two nitrogens resonated lower field. A two protons triplet (J 6.3 Hz) resonating at 8 4.64 was assigned to the methylene
integrating to two protons resonated lower field at 5 5.25 was assigned to the methylene group attached to the positively charged iminium nitrogen atom which caused the deshielding. The indazole proton at position-3 resonated as a very low field singlet at 5 1 0 .0 2 as a result of considerable deshielding by the positively charged
iminium group. The four aromatic protons resonated discretely as four peaks. The triplet (J 8.1 Hz) resonating at 8 7.45 was assigned to H-5 which was relatively the
most shielded of the aromatic protons as a result of being adjacent to the electron releasing tertiary nitrogen atom. The H-4 and H-7 protons resonated as two sets of doublet (J 8.1 Hz) at 8 7.65 and 8 8.25 respectively whilst the H-6 proton resonated as
a triplet at 8 7.80.
Having established the reaction conditions for constructing the six-membered heterocyclic ring we then proceeded to try it out on the 4,6-dimethoxy-3-methyl-1H- indazole 47. Reaction of the indazole 47 with 1,4-dibromobutane in 1,4-dioxane at
150°C for 48 h produced the desired product 79 in 78% yield (scheme 19).
Scheme 19 MeO © Br Dioxane MeO OMe Me MeO" 47
The structure of 79 was established from its !H nmr spectrum recorded in deuterochloroform containing a few drops of dimethyl sulphoxide in which the various peaks were broad compared to the spectrum of indazole 78. Some characteristic features of the spectrum were that the internal methylene protons resonated as a four proton multiplet at 8 2.10 and the methyl group at position-3
resonated as a three proton singlet at 8 2.85. The methylene group attached to the
nitrogen atom at position-1 of the indazole resonated as a multiplet at 8 4.28 whilst
the other methylene group attached to the iminium nitrogen atom resonated as a multiplet at 8 4.40. This was in contrast with the resonance position of 8 5.25 for the
methylene protons attached to the iminium nitrogen in indazole 78. This upfield shift observed in 79 for the methylene protons attached to the iminium nitrogen is
to the iminium group thus decreasing its electrophilic nature. The electrospray mass spectrum gave (M +l) ions at m/z 327/329 consistent with the molecular structure of 79.
o©
h3c
h3co
Nigellicine Nigellidine
During this research to develop methods for the synthesis of Nigellicine 1 the
synthesis of Nigellidine 4 for which the ketone 53 was a key intermediate compound, was also addressed.
The initial synthetic strategy for the synthesis of ketone 53 involved the Friedel- Crafts acylation of 1,3,5-trimethoxybenzene with 4-benzyoxybenzoyl chloride 6a. as
shown in scheme 20. The acid chloride was freshly prepared as described in chapter one of this thesis and was reacted with 1,3,5-trimethoxybenzene with 4-benzyloxy- benzoyl chloride 6a in both CS2 and DCM as reaction solvents. In all the cases the
reaction failed to give any desired product 53. In fact the product isolated from the reactions was identified from its JH NMR spectrum as 4-benzyloxybenzoic acid 5a. Literature survey suggests that one limitation for the Friedel-Crafts reaction is that it doesn't succeed on an aromatic ring that is substituted either by an amino group or by
fn
a strongly electron-withdrawing group . In this case the steric factors as well as the Scheme 20
DCM
Ph
53
reaction. The failure of the Friedel-Crafts reaction requires an alternative method for the synthesis of ketone 53 to be considered. The synthesis of the ketone 53 by oxidation of the alcohol 52 which could be made by a Grignard reaction between 2,4,6-trimethoxybenzaldehyde 50 and the Grignard reagent 51 derived from benzyl 4-bromophenyl ether as shown in scheme 21, appeared possible.
Scheme 21
The synthesis of the Grignard reagent was initially tried in dry diethyl ether but it failed to form. Even when dry THF was used as the reaction solvent, the formation of the Grignard reagent 51 required strong heating as well as a crystal of iodine as catalyst. The chiral alcohol 52 was obtained in 55% yield after purification by column chromatography as a racemic white crystalline solid. The compound 52 was
characterised spectroscopically. Some characteristic features of its ]H NMR spectrum, recorded in deuterochloroform, were the high field resonance of the OH group at 5 1.60 and the low field resonance in the aromatic region of the tertiary proton on the carbon bearing the OH group at 5 6.15-6.23. The 1,4-disubstituted aromatic ring
protons resonated as an AB system at 8 6.90 and 8 7.25. The electron ionisation mass
spectrum of 52 showed a molecular ion at m/z 380 and a fragment ion at m/z 197 was a base peak indicative of the loss of benzyloxyphenyl group.
Oxidation of 52 to the ketone 53 was initially tried with commercially available activated manganese dioxide in DCM and chloroform at different concentrations of the oxidant. All the reactions were monitored by TLC and no product formation was observed. Oxidation of the alcohol 52 by Jones reagent in acetone 0°C for a short time did not give any product 53. However, oxidation with Jones reagent for a prolonged